The matching law states that the absolute amount of each form of behavior is proportional to the relative amount of reinforcement associated with it. While the law has been confirmed, at least approximately, in numerous experiments, a number of potentially significant disconfirmations have been reported. The main objective of the work proposed is to test the limits of generality of the law and, possibly, to revise it. The law of effect appears to be best considered a principle of motivation, rather than of learning. Most of the variables to be studied are consequently motivational, by which is meant parameters of drive and incentive. However, some of the work deals with the relationship between the law of effect and the process of association.